Moral luck

Moral luck is the irreducible contingency that unsettles any attempt to judge an ethical act by a single, stable measure. An act’s value cannot be secured by pure intention alone, yet its consequences do not provide a final verdict from outside history: they emerge in a world the agent cannot fully control, and the person who later judges the act has been changed by having lived through it. The concept therefore exposes the strain between principle, outcome, and the shifting standpoint from which a life is assessed.

Beyond Intention and Consequence

The discussion places Bernard Williams between two familiar ethical positions. Against Kant, one cannot treat the purity of the will as a complete measure of moral value, independent of what an act brings about. Against a consequentialist calculation, however, moral judgment cannot be reduced to a common currency that converts different actions and experiences into comparable quantities, whether that currency is the moral Law or utility. Williams questions the prospect of a universal medium for settling every moral case: the circumstances in which an action takes place matter, and their significance cannot always be abstracted away.

This position also resists the easy assurance that good consequences vindicate an act. A reference to the greater good may justify harm done to particular people, even where the justification has the appearance of rational calculation. But refusing a morally distasteful action is not automatically a secure alternative. Williams identifies the risk of moral self-indulgence in the suspicion that the agent is less concerned for other people than with preserving the image of himself as someone who cares about them. The discomfort of an action, then, is not by itself a moral argument either for performing it or for refusing it.

Williams’s objection reaches beyond the familiar dispute between duty and utility to the idea that a person can organize a life according to a rationally settled plan. A life cannot be evaluated against a complete set of preferences fixed in advance, because what a person does—and the kind of life those actions form—helps condition the desires and judgments that person will have later. The standard of assessment is therefore not independent of the history it is supposed to assess.

The Retrospective Self

This dependence makes time central to moral judgment. Deliberation begins from the agent’s present position, with the knowledge and limits available there. Retrospective assessment takes place later, from a standpoint informed by what happened and by the person the intervening choices have helped produce. The later self is not simply an impartial observer waiting to score the earlier self against a pre-existing scale; the earlier life has contributed to the later judge’s desires, commitments, and sense of what matters.

For that reason, a person cannot guarantee in advance what the future assessment will be, either in its factual circumstances or in its point of view. Even the deepest regrets cannot be specified with certainty beforehand. An action may alter the very outlook from which it will later be judged, so the agent cannot fully know what the action will mean to the person who remembers it. Moral luck names this exposure to contingency, not merely an accidental difference between an intended act and its outcome, but the impossibility of securing a final moral perspective outside the unfolding of a life.

This temporal structure also complicates the claim that people are responsible to themselves as one person across time. If a rational individual is told to choose so as never to blame himself, whatever happens, the advice assumes a stable future judge whose standards can already be consulted. Williams’s argument unsettles that assumption: the choices being evaluated help make the evaluator. There is no untouched standpoint from which every possible version of one’s life can be compared and ranked.

Kant, Hegel, and the Opacity of Action

The argument does not simply replace Kantian intention with confidence in outcomes. Kant’s emphasis on autonomy can be read as an acknowledgment of a real limitation: people act in situations whose consequences remain opaque to them. Since no agent can control the full course or reception of an act, sincere intention retains moral importance. The limit of consequence-based judgment is not that consequences never matter, but that they cannot be mastered in advance or made into a completely reliable standard.

Nor does Hegelian trust in ethical substance settle the problem. If the social world is treated as benevolent and the truth of an act is identified with the consequences inscribed in that world, the account assumes a harmony between individual intention and social reality. But social substance may be unjust: it can turn an act into the opposite of what its agent meant to accomplish. In that situation, the mismatch between intention and result cannot simply be charged to the agent. Hegel’s acknowledgment that the mob may revolt against social substance marks the possibility that the social order itself, rather than the individual’s refusal to identify with it, is at fault.

Moral luck thus opens a gap between intending, acting, and being judged. Kantian purity of intention risks making the agent’s inner conviction the sole criterion; confidence in ethical substance risks treating the actual social result as morally trustworthy. Williams’s emphasis on contingency refuses both guarantees. The circumstances of an action and the history that follows it matter, but neither can be converted into a secure, universal measure that erases the agent’s limited position.

The Leap of Faith

The account finally turns the problem back upon Kant. Although Kant appears to exclude moral luck by refusing to make consequences the criterion of moral worth, morality cannot be practiced by confining oneself to inner intention and dismissing what actually happens. Moral action requires an expectation that intention and the objective structure of reality are not wholly alien to one another. In Kant’s framework, the postulates of pure practical reason support that necessary affinity; in the book’s formulation, moral agency therefore involves a leap of faith and a fundamental trust in the friendliness of reality.

That trust does not remove contingency or promise that every good intention will receive a good result. It identifies what ethical action must presuppose if it is to remain more than private purity: that the world might answer, at least in some measure, to moral commitment. The danger is evident in both directions. Without regard for consequences, intention can become self-enclosed; with complete faith in consequences, the agent may mistake an unjust social order for a moral verdict. Moral luck persists in the unresolved interval between the two.

The discussion gives this interval an artistic emblem in the trio “Soave sia il vento” from Mozart’s Così fan tutte. Its appeal asks the elements to respond benignly to human desires. The image is not proof that the world is benevolent; it makes audible the hope that practical morality must risk. To act sincerely is not to know what the world will do with one’s action. It is to proceed without that knowledge, while relying on a harmony that cannot be guaranteed in advance.